QuantConnect MCP Server

Instructions

Implementation Reference

• quantconnect_mcp/src/tools/backtest_tools.py:454-564 (handler)

  The main handler function for the 'read_backtest_insights' tool. It validates input parameters, authenticates with QuantConnect API, makes a POST request to 'backtests/read/insights' endpoint to fetch insights data within the specified range, and returns formatted success or error response.

  @mcp.tool()
  async def read_backtest_insights(
      project_id: int, backtest_id: str, start: int = 0, end: int = 100
  ) -> Dict[str, Any]:
      """
      Read insights from a backtest.
  
      Args:
          project_id: ID of the project containing the backtest
          backtest_id: ID of the backtest to read insights from
          start: Starting index of insights to fetch (default: 0)
          end: Last index of insights to fetch (default: 100, max range: 100)
  
      Returns:
          Dictionary containing insights data and total count
      """
      auth = get_auth_instance()
      if auth is None:
          return {
              "status": "error",
              "error": "QuantConnect authentication not configured. Use configure_auth() first.",
          }
  
      # Validate range
      if end - start > 100:
          return {
              "status": "error",
              "error": "Range too large: end - start must be less than or equal to 100",
          }
  
      if start < 0 or end < 0:
          return {
              "status": "error",
              "error": "Start and end indices must be non-negative",
          }
  
      if start >= end:
          return {
              "status": "error",
              "error": "Start index must be less than end index",
          }
  
      try:
          # Prepare request data
          request_data = {
              "projectId": project_id,
              "backtestId": backtest_id,
              "start": start,
              "end": end,
          }
  
          # Make API request
          response = await auth.make_authenticated_request(
              endpoint="backtests/read/insights", method="POST", json=request_data
          )
  
          # Parse response
          if response.status_code == 200:
              data = response.json()
  
              if data.get("success", False):
                  insights = data.get("insights", [])
                  length = data.get("length", 0)
  
                  return {
                      "status": "success",
                      "project_id": project_id,
                      "backtest_id": backtest_id,
                      "start": start,
                      "end": end,
                      "insights": insights,
                      "length": length,
                      "message": f"Successfully retrieved {length} insights from backtest {backtest_id} (range: {start}-{end})",
                  }
              else:
                  # API returned success=false
                  errors = data.get("errors", ["Unknown error"])
                  return {
                      "status": "error",
                      "error": "Failed to read backtest insights",
                      "details": errors,
                      "project_id": project_id,
                      "backtest_id": backtest_id,
                  }
  
          elif response.status_code == 401:
              return {
                  "status": "error",
                  "error": "Authentication failed. Check your credentials and ensure they haven't expired.",
              }
  
          else:
              return {
                  "status": "error",
                  "error": f"API request failed with status {response.status_code}",
                  "response_text": (
                      response.text[:500]
                      if hasattr(response, "text")
                      else "No response text"
                  ),
              }
  
      except Exception as e:
          return {
              "status": "error",
              "error": f"Failed to read backtest insights: {str(e)}",
              "project_id": project_id,
              "backtest_id": backtest_id,
              "start": start,
              "end": end,
          }

• quantconnect_mcp/src/server.py:77-77 (registration)

  Registration call that invokes register_backtest_tools(mcp), which defines and registers the read_backtest_insights tool (and other backtest tools) with the FastMCP server instance.

  register_backtest_tools(mcp)

• quantconnect_mcp/main.py:50-50 (registration)

  Alternative entry point registration call that registers the backtest tools including read_backtest_insights.

  register_backtest_tools(mcp)